Content-based image retrieval: Concept and current practices

Abstract

Research in content-based image retrieval (CBIR) today is a lively discipline and expanding in breadth. Content-based image retrieval is a computer vision technique to the image retrieval problem of searching for digital images in huge databases. Weather forecasting, data mining, remote sensing, medical imaging, education, crime prevention and management of earth resources are a few domains where content-based image retrieval technique is in huge demand. To improve visual similarity search and image retrieval process in content-based image retrieval many studies have been conducted and methods developed in recent years, but there are a few issues that need to be addressed. This paper explores the current practices in content-based image retrieval and their effectiveness.